面向分频海上风电系统的模块化多电平矩阵变换器混合建模与控制

Hybrid Modeling and Control of Modular Multilevel Matrix Converter for Offshore Fractional Frequency Transmission System

分频输电(fractional frequency transmission system,FFTS)结合高压交流和高压直流输电的优势,是极具发展前景的大规模、中远海风电输送方案。模块化多电平矩阵变换器(modular multilevel matrix converter,M^(3)C)以其控制性能好、易于冗余扩展等优点,在海上风电FFTS中备受关注。然而,M^(3)C-FFTS中不同频率的输入和输出直接耦合,给系统建模与控制带来了挑战。为解决此问题,该文提出一种适用于M^(3)C的混合建模方法及相应的控制策略。对各子换流器的3个桥臂进行差模–共模分解,实现输入–输出解耦;对各子换流器桥臂功率进行αβ0建模,并分析桥臂功率低频分量与差模电流基波分量的约束关系。在此基础上,提出一种新型的M^(3)C-FFTS系统控制策略,通过构造差模电流的基波正序有功、负序有功和无功分量实现子模块电容电压平衡控制。所构造差模电流仅包含输入工频分量,显著降低控制策略的复杂度。最后,在220k V/400MW M^(3)C-FFTS中验证所提建模方法与控制策略的有效性。

Fractional frequency transmission system(FFTS) combines the advantages of HVAC and HVDC transmission system, thus it is a very promising solution for large-scale offshore wind power transmission. Modular multilevel matrix converter(M^(3)C) which enjoys advantages of high power quality and easy scalability has attracted much attention in offshore FFTS. However, the input side and output side of M^(3)C are directly coupled with different frequencies,which brings challenges to its modeling and control. To address this issue, a hybrid modeling method and the corresponding control strategy were proposed for M^(3)C in this paper. The proposed method applied differential-common-mode modeling to the subconverters in order to realize the decoupling of input and output. Besides, αβ 0 modeling was applied to the arm power of each subconverter, which revealed the relationship between the fundamental component of the differential-mode current and the low-frequency component of the arm power.Based on the hybrid modeling, a novel system control strategy of M^(3)C-FFTS was proposed, in which the positive sequence active component, negative sequence active and reactive components of the differential-mode current were utilized to realize capacitor voltage balancing. The constructed differential-mode current only contained the input frequency components, which significantly simplified the design of the controller. The effectiveness of the proposed modeling and control strategy was verified by a 220kV, 400MW M^(3)C system implemented in Matlab.